Adaptive Molecular Evolution

1
Adaptive Molecular Evolution

• Nonsynonymous
• vs
• Synonymous

2
Reading for today

• Li and Graur chapter (PDF on website)
• Evolutionary EST paper (PDF on website)
• Page and Holmes pp. 231 - 243

3
Predictions of neutral theory

• There is an inverse correlation between rate of
  substitution and degree of functional constraint.
• Patterns of base composition and codon usage
  reflect mutational rather than selective
  pressures.
• There is a constant rate of molecular evolution.
• The level of within species variation is the
  product of population size and mutation rate and
  is correlated with levels between species.

4
The neutral theory of molecular evolution
5
Well accepted ruleEvolutionarily
conservation of genes and regions implies
functional importance
6
(No Transcript)
7
What molecular changes are different between
species?
8
Rodent x Human 1880 Orthologous Sequence Pairs
(4 of genes)
1000

• Genes involved in immune response
• Genes involved in olfaction
• Genes involved in reproduction
• Genes implicated in human disease?

948
800
NUMBER OF PAIRS
600
483
400
200
238
138
57
12
2
10 20 30 40 50
60 70
Amino Acid Sequence Divergence
Makalowski Bogusti, PNAS 95, 9407 (1998)
9
Potential causes of rapid evolution

• Lack of constraint Selectively neutral
  evolution
• Adaptive value for change Positive
  Darwinian selection

Compare cDNA sequences.
10
2 types of changes in codons
Synonymous silent change (amino acid stays the
same) Nonsynonymous replacement change (changes
amino acid)
Val GTC
Nonsynonymous Change
Synonymous Change
Val GTG
Ala GCC

11
Rates of synonymous changes is similar to
pseudogenes
Synonymous changes
12
(No Transcript)
13
Synonymous and nonsynonymous sites are both in
coding regions. Synonymous sites are considered
selectively neutral. Therefore, we can use
synonymous sites as a ruler for nonsynonymous
substitutions. When nonsynonymous changes
exceeds synonymous changes, infer positive
selection.
There are more nonsynonymous than synonymous
sites in coding DNA
Prot. 1 Ile Cys Ile Lys Ala Leu Val Leu Thr DNA
1 ATA TGT ATA AAG CGA GTC CTG TTA ACA DNA2
ATA TGT ATA AAG CGA GTC CTG TTA ACA Prot. 2
Ile Cys Ile Lys Ala Leu Val Leu Thr
14
dN nonsynonymous substitutions/
nonsynonymous sites dS synonymous
substitutions/ synonymous sites
Test for selection by comparing dN and dS dN /dS
1 Neutral evolution dN /dS lt 1 Purifying
selection dN /dS gt 1 Positive selection
The dN/dS ratio (?) measures the selective
pressure
15
Multiple methods for calculating dN /dS

• Counting methods
• Nei and Gojobori
• Li et al.
• Maximum likelihood methods (model of codon
  evolution)
• Muse and Gaut
• Neilsen and Yang

16
Codon degeneracy

• Non-degenerate
• All mutations produce nonsynonymous change
• Two-fold degenerate
• one of the three possible changes is synonymous
• Four-fold degenerate
• all mutations produce synonymous change

17
When counting sites

• Non-degenerate (1)
• nonsynonymous
• Two fold degenerate (2)
• 1/3 synonymous and 2/3 nonsynonymous
• Four fold degenerate (4)
• synonymous

Note Three fold degenerate treated as two-fold.
18
Example Degeneracy 1 Asp Thr Ala Val Sequence
1 GAC ACA GCG GTT
How many synonymous sites in sequence 1? First,
assign degeneracy to each codon position.
19
Example Degeneracy 1 1 Asp Thr Ala Val Sequenc
e 1 GAC ACA GCG GTT
20
Example Degeneracy 1 11 Asp Thr Ala Val Sequen
ce 1 GAC ACA GCG GTT
21
Example Degeneracy 1 112 Asp Thr Ala Val Sequ
ence 1 GAC ACA GCG GTT
22
Example Degeneracy 1 112 1 Asp Thr Ala Val Seq
uence 1 GAC ACA GCG GTT
23
Example Degeneracy 1 112 11 Asp Thr Ala Val Se
quence 1 GAC ACA GCG GTT
24
Example Degeneracy 1 112 114 Asp Thr Ala Val
Sequence 1 GAC ACA GCG GTT
25
Example Degeneracy 1 112 114 1 Asp Thr Ala Val
Sequence 1 GAC ACA GCG GTT
26
Example Degeneracy 1 112 114 11 Asp Thr Ala Va
l Sequence 1 GAC ACA GCG GTT
27
Example Degeneracy 1 112 114 114 Asp Thr Ala
Val Sequence 1 GAC ACA GCG GTT
28
Example Degeneracy 1 112 114 114 1 Asp Thr Ala
Val Sequence 1 GAC ACA GCG GTT
29
Example Degeneracy 1 112 114 114 11 Asp Thr Al
a Val Sequence 1 GAC ACA GCG GTT
30
Example Degeneracy 1 112 114 114 114 Asp Thr A
la Val Sequence 1 GAC ACA GCG GTT
31
Example Degeneracy 1 112 114 114 114 Asp Thr A
la Val Sequence 1 GAC ACA GCG GTT
How many nonsynonymous sites in sequence 1?
32
Example Degeneracy 1 112 114 114 114 Asp Thr A
la Val Sequence 1 GAC ACA GCG GTT
How many nonsynonymous sites in sequence 1? 8
nondegenerate sites 1 two fold degenerate site
8.66 nonsynonymous sites
33
Example Degeneracy 1 112 114 114 114 Asp Thr A
la Val Sequence 1 GAC ACA GCG GTT
How many synonymous sites in sequence 1?
34
Example Degeneracy 1 112 114 114 114 Asp Thr A
la Val Sequence 1 GAC ACA GCG GTT
How many synonymous sites in sequence 1? 3 four
fould degenerate sites, 1 two fold 3.33
synonymous sites.
35
Example Degeneracy 1 112 114 114 114 Asp Thr A
la Val Sequence 1 GAC ACA GCG GTT Sequence 2
GCC ACT TCG GTT Ala Thr Ser Val Degeneracy
2 114 114 114 114
Sequence 2 has 8 nonsynonymous sites and 4
synonymous sites. For this comparison, we
average number from both sequences. Nonsynonymous
sites (8.66 8)/2 8.33 Synonymous sites
(3.33 4) 3.67
36
Example Degeneracy 1 112 114 114 114 Asp Thr A
la Val Sequence 1 GAC ACA GCG GTT Sequence 2
GCC ACT TCG GTT Ala Thr Ser Val Degeneracy
2 114 114 114 114
There are 2 nonsynonymous changes, So dn
2/8.33 0.24 There is 1 silent change, So ds
1/3.67 0.27 dn/ds 0.23/0.27 0.88 lt 1
despite having more nonsynonymous changes.
37
Other factors can effect calculation of dN/dS

• Transition/transversion ratio
• Transitions typically more frequent
• Pathway of substitution
• Codon bias

38
Nearly all counting methods assume all pathways
are equally likely.
39
Codon Bias

• Unequal codon usage results in reduced number of
  effective codon sites.
• Ignoring codon bias leads to underestimate of ds.

40
Maximum likelihood methods incorporate models of
codon evolution / bias.
Transition
41
Problems with dn/ds for detecting selection

• Positive selection acting only on a few sites
  (binding cleft).
• Burst of positive selection followed by purifying
  selection (lineage specific events).
• Positive selection in promoter and non-coding
  regions.
• Positive selection for post-translational
  modification (glycosylation).